The present invention relates generally to an Internet protocol (IP) based system, and more particularly, to an IP based system that facilitates communication and quality of service (QoS) between remote user terminals and Web servers across parts of the Internet that are configured by an asynchronous transfer mode (ATM) network or other communication networks.
An asynchronous transfer mode network is a network that can transfer information from one or more sources to one or more destinations. The ATM network can be deployed for configuring parts of the Internet, thus comprising a communication network. The communication network itself may be composed of multiple communicating nodes (e.g., terminals, routers, servers, switches, etc.) that are interconnected to each other by physical communication links (wires, cables, optical fibers, RF channels, etc.). An ATM-equipped node transmits a signal containing a bit stream to an adjacent node via the communication link that connects the two adjacent nodes. The transmitted bit stream is organized into fixed sized packet or xe2x80x9ccellxe2x80x9dslots for carrying, e.g., 53 byte packets called cells. Illustratively, each cell has a 5 byte header for communicating control information, and a 48 byte payload for communicating a message to be conveyed between nodes. A node allocates a xe2x80x9cvirtual channelxe2x80x9d to each communication, which, amongst other things, identifies an adjacent node to which cells of the communication must be transmitted. A sequence of virtual channels of nodes on a path between a source node and a destination node identifies a virtual channel connection.
The source node transmits cells to the destination node via this sequence of virtual channels, i.e., from node to node on the path, in a bucket brigade like fashion. Prior to transmitting the information, the source node segments the information into 48 byte sized messages and appends a 5 byte header to each such message to form a cell. The source node writes a virtual address, such as a virtual channel identifier, which enables each node on the path that receives the cell to determine the outgoing virtual channel on which to transmit the cell. A destination node receiving the cells extracts the messages from the payloads and reassembles the messages (in the appropriate order) into the originally transmitted information.
It is important to note that, when communicating between Local Area Networks (LANs) (such as Ethernets) on the Internet which is configured by various links and networks including the ATM network, the data-link layer communications in LAN networks are different than those in the ATM networks. Thus, bridges are utilized to receive information from one network (e.g., LAN) transmitted according to the respective data-link layer and retransmit the information in the other network (e.g., ATM network) according to its respective data-link layer. In other words, bridges decouple the two incompatible data-link layers from each other, yet enable communication between nodes in each of the two networks.
At the network layer, all of the nodes may communicate using the same protocol, e.g., the Internet protocol (IP). Like the Ethernet protocol, each node that can serve as a source or destination node is assigned an IP identifier, or IP address. Information is transmitted from a source node to a destination node in a bit stream that is organized into packets. As previously stated, each packet has a header and a payload. The IP address of the source node is written in a source field of the packet header, and the IP address of the destination node is written in a destination field of the packet header. The data is then written in the payload. The packet is then transmitted according to the appropriate data-link layer protocol for the network (e.g., formed into Media Access Control (MAC) frames, divided into ATM cells, etc.) and then transmitted to its respective destination node. IP provides a routing function for routing a packet from node to node in a sequence of nodes until the packet arrives at its destinations using routing tables.
Such communications networks are becoming increasingly important vehicles for delivering video, audio and other data to and from remote user terminals. For example, such networks are used to support video-on-demand, near video-on-demand, and pay-per-view applications. However, problems are evident in terms of adequate bandwidth.
Typically, wideband (1-10 Mb/s) access technologies are relatively expensive and specialized (such as T1 lines), such that their use has been primarily by large institutional customers, such as large corporations, universities, and government agencies. However, a number of new network access technologies are now moving from the research labs into general availability. For example, some network providers have started to deploy hybrid fiber coax access lines, as well as cable modems. Further, trials and limited deployments of a variety of Digital Subscriber Line (xDSL) (e.g., Asymmetrical Digital Subscriber Linesxe2x80x94ADSL) technologies are also ongoing. All of this activity is intended to bring broadband networks to the mass market. The ATM network, as previously described, is intended to xe2x80x9cbridge the gapxe2x80x9d in providing wideband transmission rates to a remote user in an Internet environment.
A conventional IP based system is shown in FIG. 1. System 10 includes groups 12a, 12b through 12N of remote user terminals 15, where each group is part of an Ethernet or other LAN system. Each group 12a, 12b through 12N is connected to a respective Ethernet bridge, illustratively an Asymmetrical Digital Subscriber Line (ADSL) termination unitxe2x80x94remote side (as opposed to a central office side) ATU-R 18a, 18b through 18N, respectively, for providing an Ethernet bridge between the LAN system and the respective default IP router. ATU-Rs 18a, 18b, 18c through 18N are also connected to respective telephones 16a, 16b, 16c through 16N.
System 10 further includes a plurality of ADSL based Digital Subscriber Line Access Multiplexors (DSLAM) 20a through 20N, each of which is connected to a plurality of ATURs. Each DSLAM provides basic transport and multiplexing functions between each respective ATU-R 18 and an ATM switch 22. ATM switch 22 is further connected to a plurality of Web servers 28a through 28N, IP edge routers 26a through 26N and IP backbone router 24 for connection to the Internet. The function of each will be described below.
First, we describe the ADSL technology. ADSL was motivated by the goal of achieving wideband transmission rates over existing copper loops. The concept has achieved a growing acceptance by the telecommunications industry and resulted in a standardization effort.
The main idea behind ADSL is that overlapping parts of the spectrum should be present only for signals that are propagating in the same direction within a single bundle of copper wire pairs. This approach reduces the effect of near-end and far-end crosstalk, and hence makes wideband transmission rates feasible for reasonably large loop lengths (up to 18,000 ft). Several characteristics of traditional ADSL systems are specifically tailored to their intended application in residential local loops. These include the asymmetry of bandwidth and the support for life-line telephony as an inseparable component. Note that ADSL technology is suitable for Internet access service. For example, a 10 baseT (10 bT) interface is located at the ATU-R for personal computer protocols.
A DSLAM is utilized in the IP based system to support ATM bearer service for IP applications. In an ATM bearer service scenario, the interface on the user side is configured as an ATM user to network interface over an ADSL. On the ATM network (trunk) side, the interface is configured as a network to network interface over a synchronous optical network (SONET) transport. Typically, DSLAMs are located in the central office; however they can also be remote such that they are connected over significant distances by single mode fiber links.
DSLAMs support life-line telephony and an asymmetric high/low speed data channel. The DSLAM upstream data channels typically support data rates of 9.6-156 kb/s. The downstream bit rate is either fixed or distance-dependent. For some early commercial products, a 2.3 Mb/s rate was specified for distances of no more than 3 km, and about 4 Mb/s can be supported over loops shorter than 4 k m.
The number of subscriber lines per OC-3c trunk is a function of the Quality of Service (QoS), which is required from the ATM bearer service to support the target set of applications. Each OC-3c interface on the ATM network side of the DSLAM provides a line rate of 155.52 Mb/s, giving an effective ATM cell rate close to 150 Mb/s, which leaves an effective bit rate of about 135 Mb/s. If, for example, a non-blocking streaming service is supported for video applications of a constant bit rate of 2.5 Mb/s, then a DSLAM can be configured to support 48 ADSL lines (2.5 Mb/sxc3x9748=120 Mb/s). Note that DSLAMs have been designed to provide access to a broadband backbone, characterized by the ATM transport service on the top of the SONET broadband digital hierarchy.
As described above, ADSL-based DSLAMs provide subscriber line multiplexing functionality for Internet Protocol based system. As shown in FIG. 1, DSLAMs 20 crossconnect the user terminals 15 or Ethernet 12 to the IP edge routers 26, via ATM switch 22.
A conventional router suffers performance degradation due to processing overloads since it typically support a large population of users. Accordingly, it is common for the router administrators to turn off some important functions (e.g., packet filtering, RSVP, etc.) to improve the router performance.
FIG. 2 illustrates the protocol stack for the system of FIG. 1. As shown, protocol stack 30 from the user terminals (or PCs) 15 includes a network layer, a data link layer and the physical or PHY layer. The network layer comprises the IP which includes the destination and source addresses, the data link layer comprises the link layer control (LLC) and media access control (MAC) which includes information pertaining to when to transmit and how to construct the frame, and the PHY comprises the 10bT which represents a 10 MHz NRZI signal.
Protocol stack 32 from the ATU-Rs 18 consists of ADSL layer (physical and data link), ATM, and ATM adoption layer (AAL) on the DSLAM side and 10bT and MAC layer on the user side. Protocol Stack 38 at the IP Router includes the SONET, ATM, AAL, and LLC on the ADSL (user) side and SONET, ATM, and AAL on the network side. Protocol stack 34 from the DSLAM comprises ATM and ADSL on the user side and ATM and SONET on the network side. The ATM switch 22 maintains the layer protocols. Protocol stack 38 provides simulation of the Ethernet bridge on the user side by injecting the LLC-layer.
An example of the operation of the conventional Internet based system 10 is as follows. Consider that Web server 28a wishes to communicate with user terminal 15 that resides on LAN 12b. Now, depending on which IP edge routers are assigned to Web server 28a and to this terminal, there will be at least two passes of the data packets through the ATM switch 22. Consider a best case scenario in which Web server 28a and terminal 15 on LAN 12b are both assigned to IP edge router 26N. Thus, the minimum path between Web server 28a and terminal 15 on LAN 12b must include a link from Web server 28a to IP edge router 26N through ATM switch 22. Then a link must be established from IP edge router 26N, through ATM switch 22, to LAN 12b, via DSLAM 20a. This path passes twice through ATM switch 22, thus contributing to traffic. Of course, if the Web server and terminal were assigned to different IP edge routers, then a minimum of three passes through ATM switch 22 are necessary, since an additional link would also have to be established from the IP edge router assigned to the terminal and the IP edge router assigned to the Web server via the ATM switch 22.
Note that under the architecture of FIG. 1, the DSLAMs functionality are limited to providing transport and multiplexing functions. The IP edge routers, in addition to the IP routing mechanism, provide the Ethernet bridging capability in order to be able to address the user""s LANs. Accordingly, the system of FIG. 1 suffers from a number of performance deficiencies.
Specifically, the IP edge routers become bottlenecks due to the large amount of processing that they have to perform on in-bound and on out-bound packets. It becomes tw difficult to support QoS in the routers since the bandwidth and processing resources are in a deficit. Consider if a user requires a stream of video information to be transmitted from a Web server, then this stream needs to follow a path from the Web server to an appropriate edge router, take one or more hops between the Web server""s edge router and the edge router to which the user is assigned, and then follow a path from the edge router assigned to the user to the user terminal. Only then will the stream of video information reach the user""s LAN.
Another disadvantage is that the Address Resolution Protocol (ARP) mechanism requires the router to broadcast the address to be resolved on the LANs. Thus, since the router simulates the LAN bridge, the ARP must be transmitted over the ATM network. Accordingly, additional traffic is added on the ATM network.
It is therefore an object of the present invention to overcome the deficiencies evident in the prior art.
An aspect of our present invention is an Internet protocol based system and method for facilitating communication and improving communications between remote user terminals and Web servers across parts of the Internet configured by a communications network, such as an ATM network.
The system includes a plurality of LAN networks, such as Ethernet LANs, each comprising a plurality of user terminals or PCs. The system further includes a network switch, illustratively an ATM switch, and at least one digital subscriber line (xDSL) access router, each connected between a corresponding LAN and the network switch. The xDSL be (e.g. ADSL) access routers function both as a router and a digital subscriber line multiplexor. Thus, each user terminal can communicate directly with its default router, and vice-versa, obviating the requirement of communicating with its default router via the ATM switch and reducing traffic in the ATM network.
In addition, the IP based system further comprises at least one bridge, such as an ADSL termination unit (ATU-R), each coupled between a respective LAN and a respective xDSL access router. Further, the IP based system further comprises an IP backbone router for connecting the ATM-configured network to the rest of the Internet.
Because each xDSL access router of our inventive IP based system supports a relatively small number of customers (from a few tens to a few hundreds), vis-a-vis an IP edge router, the xDSL access router can process traffic more efficiently and with far less risk of overload. In contrast, conventional routers suffer performance degradation due to processing overloads since they support much greater populations of users. Accordingly, it is common for the router administrators to turn off some important functions (e.g., packet filtering, RSVP, etc.) to improve the router performance.
In special situations, when uncompromised QoS is required (e.g., for Video On Demand Applications) the xDSL access router could serve as the default router for specialized Web servers, thus allowing direct connections from user terminals to the Web servers for better support of the IP QoS by utilizing any kind of the QoS mechanism (e.g. based on RSVP protocol) which typical servers do not utilize due to traffic constraints.